AMCPNet: Attention Mechanism and Contrast Perception-Based Algorithm for X-Ray Image Detection

X-ray Image Detection is an important detection technology in the field of public safety. In scenarios with target occlusion and complex backgrounds, which cause issues with uneven granularity of X-ray image features, this paper proposes an X-ray image detection method that integrates attention mechanisms and contrastive perception. The proposed method achieved Precision, Recall, and mAP scores of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$0.98,0.95$</tex>, and 0.97, respectively, on the PIDray dataset. The experimental results show that the proposed AMCPNet has significant performance compared with the most advanced methods, and achieves excellent results in various challenging scenarios as well as complex details.